Cerebrovascular accident, a disease commonly known as "stroke", remains the third leading cause of death, and probably constitutes the single largest category of long term disability in this country. In spite of current medical knowledge and available treatments, a major central nervous system vascular occlusion is quickly attended by irreversible damage to the affected brain region(s). A "completed stroke" is manifest by a fixed and permanent neurological deficit. Millions of dollars have been expended in stroke research and care by Federal and private agencies without a single substantial gain in our present chemotherapeutic abilities for a completed stroke.
On a clinical level, once vascular flow in any portion of the central nervous system has ceased for longer than a few minutes, a permanent "stroke" invariably follows. It is not currently possible to recover substantial neural function with clinical ischemia of 5-7 minutes duration. An exquisite neuronal sensitivity to oxygen deprivation has been blamed for this ultra-short stroke irreversibility. Neurons do indeed have meager metabolic storage and are unable to meet energy needs by anerobic means. Well accepted concepts hold that such permissible cerebral ischemia times are critical and neurons must quickly be resupplied or metabolic infarction will result. While clinically true, recent laboratory investigations have addressed the problems of ischemic vascular and neuronal reactions separately with considerably different results. Recently reported studies indicate neurons are not as sensitive as previously believed. Indeed, it has been suggested that neurons can withstand global ischemia for 1 hour or longer. K. A. Hossman, P. Kleihues, Arch. Neurol. 29, 375-389 (1973). If the clinical and experimental observations are to be reconciled, one hypothesis is tha long-term damage results from vascular rather than neuronal sensitivity to oxygen deprivation. It is known that secondary reactive changes appear within the microcirculation after sufficient stagnation. A. Ames III, R. L. Wright, M. Kowada, J. M. Thurston. G. Majno, Am. J. Pathol. 52, 437-448, (1968). J. Chiang, M. Kowada, A. Ames III, Am. J. Pathol. 52, 455-476 (1968). E. G. Fischer, Arch Neurol. 29, 361-366, (1973). E. G. Fischer, A. Ames III, E. T. Hedly-Whyte, S. O'Gorman, Stroke 8, 36-39, (1977). Even if blood is represented to the local tree, the small vessels do not completely reopen. Under these circumstances ischemic, though potentially recoverable, neurons may be lethalized because they are not adequately resupplied with blood within their metabolically tolerable limits. This concept shifts the basic fault in stroke from "ultrasensitive" neurons to a protracted blood flood failure. Nonetheless, a long felt need exists to prevent permanent damage and/or reverse neurologic deficits resulting from interrupted vascular flow.
One experimental approach which has been used to investigate the effects of stroke on neurologic tissue is the perfusion of fluids of known composition through ventriculocisternal spaces. For example, E. Fritschka, J. L. Ferguson and J. J. Spitzer have reported increases in free fatty acid turnover in cerebral spinal fluid during hypotension in dogs. According to the Fritschka technique, a "mock" cerebral spinal fluid containing radio-labelled palmitate was perfused from the lateral ventricle to the cisterna magna of conscious dogs. Arteriovenous glucose and fatty acid concentrations, and "mock" CSF fatty acid concentrations were monitored over a period of 6 hours of perfusion. Estimates of the amount of palmitate recovered from the cisternal effluent and cerebral venous blood lead to the conclusion that a sizeable fraction of free fatty acids may be taken up by tissues "in the vicinity of the CSF space". See Fritschka et al, "Increased Free Fatty Acid Turnover in CSF During Hypotension in Dogs", American Journal of Physiology, 232:H802-H807. In "Bulk Flow and Diffusion in the Cerebral Spinal Fluid System of the Goat", by Heise, Held, and Pappenheimer, a ventriculo cisternal perfusion method was used on chronically prepared, unanaesthized goats. Measurements were made of steady-state rates at which inulin, fructose, creatinine, urea, potassium, sodium, and labelled water were removed from perfusion fluid at various hydrostatic and osmotic pressures. The subject perfusions were carried out on female goats provided with implanted ventricular and cisternal guide tubes or cannulas. Each clearance period involved perfusion of 70-120 mills of fluid through the ventricular cisternal system. Inflow rate was maintained constant in the range of 1.50-2.00 ml-min, and outflow was measured continuously. The data obtained was used to investigate the effects of hydrostatic pressure on inulin clearance, rate of formation of CSF, and the permeability of the ventricular system, particularly as compared with that of the toad bladder. This ventriculo cisternal perfusion method was first reported by Pappenheimer, Heise, Jordan and Downer in "Perfusion of the Cerebral Ventricular in Unanaestheized Goats", American Journal of Physiology, Vol. 203, pp. 763-774 (1962). Pappenheimer et al reported that goals are anatomically and tempermentally suited for ventricular cisternal perfusions and can tolerate such perfusions for many hours without showing signs of discomfort. The volume of the ventricular system and rate of production of CSF are at least double corresponding values reported for large dogs, and the thickness of the goat occipital bone and its shape facilitates retrograde placement of cannulas through the occipital bone into or above the cisterna magna without interfering with muscles in the neck. The goat's horns provide natural mechanical protection for the cannulas and "are almost indispensable" for operative procedures. In accordance with the Pappenheimer et al technique, guide tubes are implanted just above the dura over the cisterna magna and just above the ependymal linings of the lateral ventricules. Prior to each perfusion the cisterna and ventricle are punctured with sharp probe needles extending a few millimeters beyond the tips of the guide tubes. Alternatively, cannulas were implanted in the subarachnoid space over the parietal cortex, thus permitting perfusion of the entire ventriculo cisternal-subarachnoid system. Pappenheimer et al followed detailed protocols for implanting the guide tubes, and for preparing sterile, synethetic CSF. The Pappenheimer et al perfusion circuit is reported to comprise a bottle sealed with a rubber cap having two stainless steel tubes extending to the bottom of the bottle. One tube serves as a gas bubbler, the second as a liquid outlet. A third opening connects with atmosphere through a sterile cotton plug. The bottle is mounted on an indicating balance and the reservoir outflow is connected through tubing to a parastalic pump with a variable drive permitting pumping rates in the range of 0.5-5 ml/min. One pump output is lead to a male syringe jo which fits the ventricular probe needles and a second outlet on the joints connects to a strain gauge manometer. A 5 ml empty sterile syringe is placed in parallel with the output to damp pulsations of the pump. The cisternal outflow is connected to an enclosed drop counter and wing flask and the output is recorded cumulatively on a polygraph which also gives a vertical record proportional to outflow rate. Pappenheimer et al reports that perfusion ith CSF of normal composition can usually be maintained for 4-8 hours before the animal becomes resistive, and if correctly performed, the animal will show no sign of knowing when the perfusion pump is on or off. No attempt is made to regulate the temperature of fluid entering the ventricular probe, however at flow rates of 1-2 ml/min it is theorized that the fluid reaches temperature equilibrium with the brain before reaching the hypothalmus. At higher flow rates (4-6 ml/min) the animals are reported to start to shiver. In this regard, see also F. H. Sklar and D. M. Long, Neurosurgery 1, 48-56 (1977).
Over the years, many experiments have been conducted with materials possessing high oxygen-dissolving properties, many of which have been incorporated as constituents in "artificial blood". The concept of utilizing materials possessing high oxygen-dissolving properties for the maintenance of tissue respiration was first reported by Rodnight in 1954. See Rodnight, R., Biochemistry Journal, Vol. 57, p. 661. Rodnight capitalized upon the considerable oxygen solubility found in silicone oils, and sustained tissue slices by incubation in these oxygen laden oils. Approximately 12 years later, Clark reported experiments involving the total immersion of small animals in silicone oils and fluorocarbon liquids. Rats totally immersed in oxygenated silicone oil survived for one hour with no apparent ill effects, but died several hours after removal, from unknown causes. Simi1ar experiments using synthetic fluorocarbon liquids, which dissolve about 3 times more oxygen than do the silicone oils, were performed with some success. Under these conditions animals survived immersion in oxygenated synthetic fluorocarbon liquids and thereafter returned to apparent health. See Clark, L. C. Jr. and Gollon F., Science, Vol. 152, p. 1755, (1966); and Gollon, F., Clark, L. C. Jr., Alabama Journal of Medical Science, Vol. 4, p. 336, (1967). While arterial oxygenation was reported as excellent for Clark's studies in rats, coincident impairment of carbon dioxide elimination was also reported, as was pulmonary damage from breathing fluorocarbon liquids. One rat, which was observed for five days following liquid breathing, was described as being in respiratory distress and as succumbing within 15 minutes after the subcutaneous administration of hydrocortisone (50 mg), with copious loss of body fluid from the trachea. In this regard, Clark concluded:
These organic liquids should prove to be of value in studies of gas exchange in living tissues in animals. Organic liquids, since they can support respiration with oxygen at atmospheric pressure and have other unique qualities, may find use in submarine escape, undersea oxygen support facilities, and medical application. The pulmonary damage caused by the breathing of the organic liquids available at the present time remains a major complication of their use in man. Science, Vol. 152, p. 1756.
See also K. K. Tremper, R. Lapin and E. Levine, Critical Care Medicine 8:738 (1980); S. A. Gould, A. L. Rosen, L. R. Sehgal, Fed. Proc. 40:2038 (1981).
Following these observations, fluorocarbon liquids were used as an incubation medium for isolated rat hearts. See Gollon and Clark, The Physiologist, Vol. 9, p. 191, (1966). In this work, myocardial oxygen requirements were apparently well met, however these hearts did not flourish without intermittent fluorocarbon removal and washing with oxygenated, diluted blood. This phenomenon has been explained in terms of aqueous phase lack in pure fluorocarbons such that necessary ionic exchange is impeded.
More recently, considerable attention has been directed to the use of fluorocarbons as constituents of artificial blood. Sloviter, in order to overcome the problem of aqueous-metabolite fluorocarbon insolubility, made an emulsion with fluorocarbon and albumin. Sloviter's emulsion sustained the isolated rat brain by a vascular perfusion as well as did an erythrocyte suspension. See Sloviter, H. A. and Kamimoto T., Nature (London), Vol. 216, p. 458 (1967). A better emulsion was later developed comprising a detergent, "Pluronic F 68" (manufactured by the Wyandotte Chemical Corp., Wyandotte, Mich.), and fluorocarbon liquids which were properly emulsified using sonic energy. This improved emulsion permitted the replacement of most of the blood of a rat which was then reported as surviving in an atmosphere of oxygen for five to six hours. See Geyer "Survival of Rats Totally Perfused with a Fluorocarbon-Detergent Preparation", Organ Perfusion and Preservation, edited by V. C. Normen, N.Y.: Appelton-Century-Crofts, pp. 85-96 (1968), Geyer, R. P., Federation Proceedings, Vol. 29, No. 5, September-October, 1970; and Geyer, R. P. Med u Ernohn, Vol. 11, p. 256 (1970).
Experiments have also been reported wherein fluorocarbons have been used to perfuse livers. Ten hours after in vitro fluorocarbon perfusion, the isolated liver ATP; AMP; lactate/pyruvate ratio; and a number of other metabolites were found to be as good or better than livers perfused in vitro with whole blood. See Krone W., Huttner, W. B., Kampf S. C. et al., Biochemika et Biophysica Acta, Vol. 372, pp. 55-71 (1974). These detailed metabolic studies indicated that the organs perfused with 100% fluorocarbon liquid were redeemed "intact"; while only 75% of the whole blood infused organs maintained a similar d gree of metabolic integrity. The ability of fluorocarbon perfusion to maintian cellular integrity was confirmed by electron-microscopy studies. The cells had normal mitochondrial ultra structure after ten hours of fluorocarbon support, indicating the persistence of normal or adequate aerobic metabolism. In Brown and Hardison, "Fluorocarbon Sonicated as a Substitute for Erythocytes in Rat Liver Perfusion", Surgery 71, pp. 388-394 (1972) a fluorocarbon perfusate preserved organ function and integrity far better than perfusate with much lower oxygen carrying capacity, but was reported as resulting in a decreased rate of bile secretion which was probably the earliest sign of hepatic damage, tissue edema, and a reproducible rise of portal pressure over a period of 21/2 to 3 hours. Both tissue edema and rising portal pressure with fluorocarbon perfusion were associated with progressive vascular occlusion as determined histogolically. A greatly diminished perfusion of fluorocarbon at the end of experiments was documented by injection of India ink twenty minutes before the end of the perfusion. Brown and Hardison hypothesized that the fluorocarbon perfusate may react with amino acids and proteins, that the oxygen concentration in the fluorocarbon perfusate may affect the perfusion results, and that filtration of the fluorocarbon emulsion through filter paper and differing instrumentation were responsible for the apparently conflicting results in the literature. Brown and Hardison hypothesize that phagocytosis of fluorocarbon particles might completely block reticuloendotheilial cells in liver or that capillary endotheilial damage may be another reason for late fluorocarbon perfusion problems.
Fluorocarbons have also been used in experiments involving cerebral blood circulation. In Rosenblum's studies, mouse hematocrits were reduced to 10-15 by exchanging the animal's blood with a fluorocarbon solution. When the animals were respired with 100% oxygen after intravascular fluorocarbon infusions, the brains remained metabolically sound. These organs were able to reverse rising NADH levels and EEG abnormalities induced by short period nitrogen inhalation. The EEG's of fluorocarbon treated animals could be activated by the central nervous system stimulant metrazole. By these criteria, intravascular fluorocarbon does support the cerebral microcirculation and provides functions of oxygenation, metabolism and electrical activity which are normally associated with blood transport. Please refer to Rosenblum, W. I., "Fluorocarbon Emulsions and Cerebral Microcirculation", Federation Proceedings, Vol. 34, No. 6, p. 1493 (May 1975). See also S. J. Peerless, R. Ishikawa, I. G. Hunter, and M. J. Peerless, Stroke 12, pp. 558-563 (1981); B. Dirk, J. Creiglstein, H. H. Lind, H. Reiger, H. Schultz, J. of Pharm. Method 4, pp. 95-108 (1980); J. Suzuki, T. Y. Oshomoto, S. Tanaka, K. Moizoi, S. Kagawa, Current Topics 9, pp. 465-470 (1981).
As reported by Kontos et al, the marked vasodilation of small cerebral surface arteries which occurs in response to acute profound hypoxemia may be locally obviated by perfusing oxygen equilibrated fluorocarbon into the space under the cranial window. See Kontos, H. A., et al, "Role of Tissue Hypoxemia in Local Regulation of Cerebral Microcirculation", American Journal of Physiology, Vol. 363, pp. 582-591 (1978). Kontos et al described the effect perfusions with fluorocarbon with 100% oxygen as resulting from increased supplies of oxygen to the neural cells and consequent partial or complete relief of hypoxia, rather than to a local increase in the oxygen tension in the immediate environment of the vascular smooth muscle of the pial arterioles. Two other potential explanations for the observed action are also suggested in the Kontos et al article.
In 1977, Doss, Kaufman and Bicher reported an experiment wherein a fluorocarbon emulsion was used to partially replace cerebrospinal fluid, with the intention of evaluating its protective effect against acute anoxia. Doss et al, Microvascular Research 13, pp. 253-260 (1977). According to this experiment, systemic hypoxia was produce through one minute of 100% nitrogen inhalation. A bolus of oxygenated fluorocarbon placed in the cisterna magna immediately prior to nitrogen breathing increased regional cerebrospinal fluid O.sub.2 tension by a factor of 5. During the one minute experimental period, the fluorocarbon emulsion provided twice as much brain tissue oxygen as was found in saline injected controls. Doss et al found the anticipated regional tissue oxygenation decline attending nitrogen inhalation to be halved by the administration of the oxygen bearing fluorocarbon emulsion.
In spite of the above described experiments, there is yet to be reported any practical therapeutic approach to the treatment of ischemic neurologic tissue, and particularly human ischemic central nervous system tissue resulting from stroke, accident or disease.